Dan Streetman wrote: >On Tue, 26 Feb 2002, Roman Weissgaerber wrote: > >>>- I'd suggested that the bandwidth would stay scheduled until >>> after an interrupt transfer completed, there was no transfer >>> queued. >>> >>If I understand you right here, I think you lose the offset (or phase) >>information >>of the interrupt schedule of the endpoint. >>E.g. if we send single shot interrupt URBs every 10ms to a 255ms >>endpoint. Then I >>think the transfer should take place every 255 ms. >>With your model there would be a transfer every 10ms. >> > >If the driver asks for 10ms polling (even though the ep says it >expects 255ms), should the HCD really second-guess the driver...? > Iff the driver asks for 10ms polling then maybe it should get a 10ms interval. But at my example the driver just sends another URB at a 'random' point of time and this other URB gets transferred immediately and this I think could be a problem.
E.g. If you send two URBs immediatly one after the other then the second one gets queued and will be transferred 255 ms later. If you send the second URB about 2ms after the first one returns then it could be transfered immediately (= about 2ms after the first URB) and this is a little bit strange. - Roman _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
